What is this?

Detect lines in a point cloud.

Subscribing Topics

  • ~input (sensor_msgs/PointCloud2)

    Input point cloud.

  • ~input_indices (jsk_recognition_msgs/ClusterPointIndices)

    Input indices of the cluster in a point cloud.

Publishing Topics

  • ~debug/line_marker (visualization_msgs/Marker)

    Marker topic to visualize detected line.

  • ~output/inliers (jsk_recognition_msgs/ClusterPointIndices)

  • ~output/coefficients (jsk_recognition_msgs/ModelCoefficientsArray)

    Result of detection.


  • ~approximate_sync (Bool, default: false)

    Whether to allow approximate synchronization of input topics.

  • ~method_type (Int, default: 0)

    The type of sample consensus method to use.

    • 0: SAC_RANSAC

    • 1: SAC_LMEDS

    • 2: SAC_MSAC

    • 3: SAC_RRANSAC

    • 4: SAC_RMSAC

    • 5: SAC_MLESAC

    • 6: SAC_PROSAC

  • ~outlier_threshold (Double, default: 0.005)

    Outlier threshold in meters to detect plane using RANSAC.

  • ~max_iterations (Int, default: 1000)

    Maximum iteration number to detect larger plane using RANSAC.

  • ~min_indices (Int, default: 1000)

    Minimum number of points which construct a line.

  • ~min_length (Double, default: 0.1)

    Minimum length of each line in meters.

  • ~line_width (Double, default: 0.01)

    Width of line marker published to ~debug/line_marker.


roslaunch jsk_pcl_ros sample_line_segment_detector.launch